A sequence-based, deep learning model accurately predicts RNA splicing branchpoints.
Ontology highlight
ABSTRACT:
SUBMITTER: Paggi JM
PROVIDER: S-EPMC6239175 | biostudies-literature | 2018 Dec
REPOSITORIES: biostudies-literature
Publications
A sequence-based, deep learning model accurately predicts RNA splicing branchpoints.
Paggi Joseph M JM Bejerano Gill G
RNA (New York, N.Y.) 20180917 12
Experimental detection of RNA splicing branchpoints is difficult. To date, high-confidence experimental annotations exist for 18% of 3' splice sites in the human genome. We develop a deep-learning-based branchpoint predictor, LaBranchoR, which predicts a correct branchpoint for at least 75% of 3' splice sites genome-wide. Detailed analysis of cases in which our predicted branchpoint deviates from experimental data suggests a correct branchpoint is predicted in over 90% of cases. We use our predi ...[more]
Similar Datasets
Project description:Identifying subgroups of T-cell acute lymphoblastic leukemia (T-ALL) with poorer survival will significantly influence patient treatment options and improve patient survival expectations. Current efforts to unambiguously predict T-ALL survival expectations in multiple patient cohorts are lacking. A deep learning (DL)-based model was developed in order to reliably determine the prognostic staging of T-ALL patients. We used transcriptome sequencing data from TARGET to build a DL-based survival model on 265 T-ALL patients, we found that patients could be divided into two subgroups (K0 and K1) with significant differences (P < 0.0001) in survival rates. The more malignant subgroup was significantly associated with some tumor-related pathways, such as PI3K-Akt, cGMP-PKG and TGF-beta signaling pathway. DL-based model also showed good performance in a cohort of patients from our clinical center (P = 0.0248). T-ALL patients survival was successfully predicted using a DL-based model, and we hope to apply it to our clinical practice in the future.
2022-10-12 | GSE214998 | GEO
Project description:Protein-DNA and protein-RNA interactions are part of many diverse and essential cellular functions and yet most of them remain to be discovered and characterized. Recent research shows that sequence-based predictors of DNA-binding residues accurately find these residues but also cross-predict many RNA-binding residues as DNA-binding, and vice versa. Most of these methods are also relatively slow, prohibiting applications on the whole-genome scale. We describe a novel sequence-based method, DRNApred, which accurately and in high-throughput predicts and discriminates between DNA- and RNA-binding residues. DRNApred was designed using a new dataset with both DNA- and RNA-binding proteins, regression that penalizes cross-predictions, and a novel two-layered architecture. DRNApred outperforms state-of-the-art predictors of DNA- or RNA-binding residues on a benchmark test dataset by substantially reducing the cross predictions and predicting arguably higher quality false positives that are located nearby the native binding residues. Moreover, it also more accurately predicts the DNA- and RNA-binding proteins. Application on the human proteome confirms that DRNApred reduces the cross predictions among the native nucleic acid binders. Also, novel putative DNA/RNA-binding proteins that it predicts share similar subcellular locations and residue charge profiles with the known native binding proteins. Webserver of DRNApred is freely available at http://biomine.cs.vcu.edu/servers/DRNApred/.
| S-EPMC5449545 | biostudies-literature
Project description:Peptide ion mobility adds an extra dimension of separation to mass spectrometry-based proteomics. The ability to accurately pre-dict peptide ion mobility would be useful to expedite assay development and to discriminate true answers in database search. There are methods to accurately predict peptide ion mobility through drift tube devices, but methods to predict mobility through high-field asymmetric waveform ion mobility (FAIMS) are underexplored. Here, we successfully model peptide ions’ FAIMS mobility using a multi-label multi-output classification scheme to account for non-normal transmission distributions. We trained two models from over 100,000 human peptide precursors: a random forest and a long-term short-term memory (LSTM) neural network. Both models had different strengths, and the ensemble average of model predictions produced higher F2 score than either model alone. Finally, we explore cases where the models make mistakes, and demonstrate predictive performance of F2=0.66 (AUROC=0.928) on a new test dataset of nearly 40,000 different E. coli peptide ions.
2025-01-16 | PXD055252 | Pride
Project description:Small non-coding RNAs (ncRNAs) are short non-coding sequences involved in gene regulation in many biological processes and diseases. The lack of a complete comprehension of their biological functionality, especially in a genome-wide scenario, has demanded new computational approaches to annotate their roles. It is widely known that secondary structure is determinant to know RNA function and machine learning based approaches have been successfully proven to predict RNA function from secondary structure information. Here we show that RNA function can be predicted with good accuracy from a lightweight representation of sequence information without the necessity of computing secondary structure features which is computationally expensive. This finding appears to go against the dogma of secondary structure being a key determinant of function in RNA. Compared to recent secondary structure based methods, the proposed solution is more robust to sequence boundary noise and reduces drastically the computational cost allowing for large data volume annotations. Scripts and datasets to reproduce the results of experiments proposed in this study are available at: https://github.com/bioinformatics-sannio/ncrna-deep.
| S-EPMC7682815 | biostudies-literature
Project description:Metabolomics holds the promise as a new technology to diagnose highly heterogeneous diseases. Conventionally, metabolomics data analysis for diagnosis is done using various statistical and machine learning based classification methods. However, it remains unknown if deep neural network, a class of increasingly popular machine learning methods, is suitable to classify metabolomics data. Here we use a cohort of 271 breast cancer tissues, 204 positive estrogen receptor (ER+), and 67 negative estrogen receptor (ER-) to test the accuracies of feed-forward networks, a deep learning (DL) framework, as well as six widely used machine learning models, namely random forest (RF), support vector machines (SVM), recursive partitioning and regression trees (RPART), linear discriminant analysis (LDA), prediction analysis for microarrays (PAM), and generalized boosted models (GBM). DL framework has the highest area under the curve (AUC) of 0.93 in classifying ER+/ER- patients, compared to the other six machine learning algorithms. Furthermore, the biological interpretation of the first hidden layer reveals eight commonly enriched significant metabolomics pathways (adjusted P-value <0.05) that cannot be discovered by other machine learning methods. Among them, protein digestion and absorption and ATP-binding cassette (ABC) transporters pathways are also confirmed in integrated analysis between metabolomics and gene expression data in these samples. In summary, deep learning method shows advantages for metabolomics based breast cancer ER status classification, with both the highest prediction accuracy (AUC = 0.93) and better revelation of disease biology. We encourage the adoption of feed-forward networks based deep learning method in the metabolomics research community for classification.
| S-EPMC5759031 | biostudies-literature
Project description:BackgroundIdentifying breast cancer patients with DNA repair pathway-related germline pathogenic variants (GPVs) is important for effectively employing systemic treatment strategies and risk-reducing interventions. However, current criteria and risk prediction models for prioritizing genetic testing among breast cancer patients do not meet the demands of clinical practice due to insufficient accuracy.MethodsThe study population comprised 3041 breast cancer patients enrolled from seven hospitals between October 2017 and 11 August 2019, who underwent germline genetic testing of 50 cancer predisposition genes (CPGs). Associations among GPVs in different CPGs and endophenotypes were evaluated using a case-control analysis. A phenotype-based GPV risk prediction model named DNA-repair Associated Breast Cancer (DrABC) was developed based on hierarchical neural network architecture and validated in an independent multicenter cohort. The predictive performance of DrABC was compared with currently used models including BRCAPRO, BOADICEA, Myriad, PENN II, and the NCCN criteria.ResultsIn total, 332 (11.3%) patients harbored GPVs in CPGs, including 134 (4.6%) in BRCA2, 131 (4.5%) in BRCA1, 33 (1.1%) in PALB2, and 37 (1.3%) in other CPGs. GPVs in CPGs were associated with distinct endophenotypes including the age at diagnosis, cancer history, family cancer history, and pathological characteristics. We developed a DrABC model to predict the risk of GPV carrier status in BRCA1/2 and other important CPGs. In predicting GPVs in BRCA1/2, the performance of DrABC (AUC = 0.79 [95% CI, 0.74-0.85], sensitivity = 82.1%, specificity = 63.1% in the independent validation cohort) was better than that of previous models (AUC range = 0.57-0.70). In predicting GPVs in any CPG, DrABC (AUC = 0.74 [95% CI, 0.69-0.79], sensitivity = 83.8%, specificity = 51.3% in the independent validation cohort) was also superior to previous models in their current versions (AUC range = 0.55-0.65). After training these previous models with the Chinese-specific dataset, DrABC still outperformed all other methods except for BOADICEA, which was the only previous model with the inclusion of pathological features. The DrABC model also showed higher sensitivity and specificity than the NCCN criteria in the multi-center validation cohort (83.8% and 51.3% vs. 78.8% and 31.2%, respectively, in predicting GPVs in any CPG). The DrABC model implementation is available online at http://gifts.bio-data.cn/ .ConclusionsBy considering the distinct endophenotypes associated with different CPGs in breast cancer patients, a phenotype-driven prediction model based on hierarchical neural network architecture was created for identification of hereditary breast cancer. The model achieved superior performance in identifying GPV carriers among Chinese breast cancer patients.
| S-EPMC8876403 | biostudies-literature
Project description:An early and accurate prediction of hepatocellular carcinoma (HCC) is beneficial for individualized treatment and follow-up of chronic hepatitis B (CHB) patients. We aimed to establish a prediction model for HCC by radiomics analysis in CHB patients and compare performance with liver stiffness measurement (LSM) and other clinical prognostic scores. Initially, 1215 patients were included and finally 434 CHB patients with 5-year follow-up were enrolled, 96.3% of them underwent liver biopsy. Deep learning radiomics analysis was performed on 2170 two-dimensional shear wave elastography (2D-SWE) and corresponding B-mode ultrasound (US) images. These high-throughput imaging features were also combined with low-dimensional serological clinical data by deep learning radiomics to establish different HCC prediction models and to overcome challenges of an unbalanced sample. The best model which is simple with high accuracy was selected. Prediction performance of the selected model was compared with LSM and other clinical prognostic scores. During 5-year follow-up, 32 (7.4%) of 434 patients developed HCC. The best prediction model was HCC-R, which included 2D-SWE and B-mode US images, sex and age. This model showed a high predictive value with areas under the receiver operating characteristic curve (AUCs) of 0.981, 0.942 and 0.900 in training, validation and testing cohorts for predicting 5-year prognosis of HCC. These predictive values were significantly higher than that of LSM (AUC: 0.676~0.784, p < 0.05) and better than that of other clinical prognostic scores (AUC: 0.544~0.869). HCC-R radiomics model based on 2D-SWE and B-mode US images, sex and age comprehensively reflected biomechanical and morphological information of patients and can accurately predict HCC occurrence; thus, this model has great value for treatment and follow-up of CHB patients.
| S-EPMC7868753 | biostudies-literature
Project description:Many cell types migrate in response to naturally generated electric fields. Furthermore, it has been suggested that the external application of an electric field may be used to intervene in and optimize natural processes such as wound healing. Precise cell guidance suitable for such optimization may rely on predictive models of cell migration, which do not generalize. Here, we present a machine learning model that can forecast directedness of cell migration given a timeseries of previous directedness and electric field values. This model is trained using time series galvanotaxis data of mammalian cranial neural crest cells obtained through time-lapse microscopy of cells cultured at 37 °C in a galvanotaxis chamber at ambient pressure. Next, we show that our modeling approach can be used for a variety of cell types and experimental conditions with very limited training data using transfer learning methods. We adapt the model to predict cell behavior for keratocytes (room temperature, ~ 18-20 °C) and keratinocytes (37 °C) under similar experimental conditions with a small dataset (~ 2-5 cells). Finally, this model can be used to perform in silico studies by simulating cell migration lines under time-varying and unseen electric fields. We demonstrate this by simulating feedback control on cell migration using a proportional-integral-derivative (PID) controller. This data-driven approach provides predictive models of cell migration that may be suitable for designing electric field based cellular control mechanisms for applications in precision medicine such as wound healing.
| S-EPMC9200721 | biostudies-literature
Project description:Accurately predicting essential genes using computational methods can greatly reduce the effort in finding them via wet experiments at both time and resource scales, and further accelerate the process of drug discovery. Several computational methods have been proposed for predicting essential genes in model organisms by integrating multiple biological data sources either via centrality measures or machine learning based methods. However, the methods aiming to predict human essential genes are still limited and the performance still need improve. In addition, most of the machine learning based essential gene prediction methods are lack of skills to handle the imbalanced learning issue inherent in the essential gene prediction problem, which might be one factor affecting their performance. We propose a deep learning based method, DeepHE, to predict human essential genes by integrating features derived from sequence data and protein-protein interaction (PPI) network. A deep learning based network embedding method is utilized to automatically learn features from PPI network. In addition, 89 sequence features were derived from DNA sequence and protein sequence for each gene. These two types of features are integrated to train a multilayer neural network. A cost-sensitive technique is used to address the imbalanced learning problem when training the deep neural network. The experimental results for predicting human essential genes show that our proposed method, DeepHE, can accurately predict human gene essentiality with an average performance of AUC higher than 94%, the area under precision-recall curve (AP) higher than 90%, and the accuracy higher than 90%. We also compare DeepHE with several widely used traditional machine learning models (SVM, Naïve Bayes, Random Forest, and Adaboost) using the same features and utilizing the same cost-sensitive technique to against the imbalanced learning issue. The experimental results show that DeepHE significantly outperforms the compared machine learning models. We have demonstrated that human essential genes can be accurately predicted by designing effective machine learning algorithm and integrating representative features captured from available biological data. The proposed deep learning framework is effective for such task.
| S-EPMC7521708 | biostudies-literature
Project description:Although research has made significant findings in the neurophysiological process behind the pupillary light reflex, the temporal prediction of the pupil diameter triggered by polychromatic or chromatic stimulus spectra is still not possible. State of the art pupil models rested in estimating a static diameter at the equilibrium-state for spectra along the Planckian locus. Neither the temporal receptor-weighting nor the spectral-dependent adaptation behaviour of the afferent pupil control path is mapped in such functions. Here we propose a deep learning-driven concept of a pupil model, which reconstructs the pupil's time course either from photometric and colourimetric or receptor-based stimulus quantities. By merging feed-forward neural networks with a biomechanical differential equation, we predict the temporal pupil light response with a mean absolute error below 0.1 mm from polychromatic (2007 [Formula: see text] 1 K, 4983 [Formula: see text] 3 K, 10,138 [Formula: see text] 22 K) and chromatic spectra (450 nm, 530 nm, 610 nm, 660 nm) at 100.01 ± 0.25 cd/m2. This non-parametric and self-learning concept could open the door to a generalized description of the pupil behaviour.
| S-EPMC7803766 | biostudies-literature